ALX4, an epigenetically down regulated tumor suppressor, inhibits breast cancer progression by interfering Wnt/β-catenin pathway

The human breast cancer cell lines MCF-7, MB-MDA-231 and T-47D were obtained from the Cell Bank of the Chinese Academy of Science (Shanghai, China) respectively. All cells were cultured in the corresponding medium (recommended by the suppliers) supplemented with 10% fetal bovine serum, and maintained at 37 °C incubator with 5% CO₂. A total of 108 primary paraffin embedded breast cancer tissue and 8 normal breast tissues of patients who had mammary anaplasty were collected from the Southwest Hospital affiliated to the Third Military Medical University. All experiments and procedures were approved by the Clinical Research Ethics Committee of the Third Military Medical University.

Total RNA was extracted from cells and tissues with Trizol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. RT-PCR and real-time quantitative PCR analyses were performed as previously described [9]. Primer sequences are listed in Additional file 1: Table S1.

WB was performed as previously described [9]. After incubation with the secondary antibody, the proteins were detected by chemiluminescence (Millipore Germany). Primary antibodies used in this study were anti-ALX4 (1:1000; Santa Cruz Biotechnology), anti-CTNNB1 (1:700; Santa Cruz Biotechnology), anti-p-CTNNB1 (1800; Santa Cruz Biotechnology), anti-GSK3β (1:800; Santa Cruz Biotechnology), anti-c-Myc (1:1000; Santa Cruz Biotechnology), anti-CCND1 (1:1000; Santa Cruz Biotechnology), anti-MMP7 (1:1000; Abcam) and anti-GAPDH (1:2000; Beyotime China). The siRNA targeting human GSK3β was purchased from Riobio (stQ0004712–1).

DNA samples were modified using the EZ DNA Methylation-Gold Kit (Zymo Research, Orange, CA, USA). The MSP and BGS were performed as previously reported [9, 18]. Primers are listed in Additional file 1: Table S1.

Cells were treated with 5-aza-2-deoxycytidine (Sigma) as previously described [9].

Briefly, 5000 cells per well were seeded in 96-well plates and transfected with pIRES2-EGFP-ALX4, ALX4-miRNA and control vectors respectively. The absorbance was determined one day after cells were plated to confirm the identical cell number of each group. Cell viability was evaluated by with Cell Proliferation Reagent MTS (Promega) according to the manufacturer’s instructions.

ALX4 or vector control-transfected cells (3 × 10⁵ cells per well) were harvested at 48-h post-transfection, and fixed in 70% ethanol overnight at 4 °C. The cells were stained with propidium iodide (BD Pharmingen, San Jose, CA, USA). A total of 30,000 cells were sorted by Accuri-C6 (BD Biosciences, Franklin Lakes, NJ, USA) and cell cycle profiles were analyzed using the Flowjo software.

A total of 142 primary Breast cancer patients who had undergone surgical resection with curative intent between 2004 and 2009, were obtained from the Southwest Hospital in Chongqing, China. The clinico-pathologic information was retrieved from the patients’ electronic medical records including age, gender, tumor size, histological grade, lymph node (negative or positive) and clinical stage (defined according to American Joint Committee on Cancer 7th edition) and follow-up information (5–10 years) for overall survival rates. This study was approved by the ethics committee of the Southwest Hospital Affiliated to Third Military Medical University, and all experiments were carried out in accordance with approved guidelines of Third Military Medical University. Informed consent was signed by all of the recruited patients. All samples from breast cancer patients were reviewed histologically by hematoxylin and eosin staining. To construct the TMA slides, two cores were taken from each representative tumor tissues (within a distance of 20 mm). The tissues were stained with hematoxylin-eosin and then reviewed histologically by at least two pathologists. Duplicate cylinders from intratumoral and peritumoral areas were obtained. Finally, the TMAs were constructed (in collaboration with Shanghai Biochip Company Ltd., Shanghai, China).

IHC staining was performed using the antibody against ALX4 as described previously [21]. The staining was evaluated for the tumor cells. The immunostaining was considered positive when ≥10% of the tumor cells being immunoreactive. A double-blind method was carried out independently by two investigators without knowing the patients’ clinical and pathological features to analyze the IHC results. Three visual fields from different areas of each specimen were chosen randomly for the IHC evaluation. ALX4 expression was scored according to staining intensity and the percentage of positive cells as previously described [22]. The percentage of positive cells was scored as follows: 0% -10% (0), 11%–30% (1), 31%–50% (2), 51%–80% (3) and 81%–100% (4). Staining intensity was scored as follows: no staining (0), week (1), moderate (2), and strong (3). Comprehensive score = staining percentage × intensity. ALX4 expression was classified as follows: ≤4 low expression, >4 high expression according to the median of 142 breast cancer patients.

The ALX4 over expression or knock down cell lines were generated as previously described [9, 21]. Briefly the PIRES2-EGFP-ALX4 vector or the pcDNA6.2-GW/EmGFP-miR vector was transfected using ViaFect transfection reagent. The stably transfected cells were screened with G418 (Calbiochem, La Jolla, CA, USA) or Blasticidin (Invitrogen Preservation). Single clone was obtained by the cylinder method. Several positive clones were confirmed by qPCR and then mixed for consequent experiments.

MDA-MB-231 and MCF-7 cells were plated in 6-well plates at 2.5 × 10⁵ cells per well. For knockdown, MDA-MB-231-ALX4 and MCF-7-ALX4 stably cells were plated in 6-well plates at 3 × 10⁵ cells per well. After culturing for 24 h, cells were transfected with ALX4 or vector control and miRNA or negative control, respectively. After 48 h of transfection, cells were collected, diluted 1:3, plated in 6-well plates and selected with 0.8 mg/ml of G418 or 2 μg/ml of Blasticidin for 14 days to establish stable clones in which the plasmids had stably integrated into genomic DNA. Surviving colonies (⩾50 cells per colony) were stained by using crystal violet (Sigma-Aldrich, St Louis, MO, USA) and counted.

For migration assay, cells (4 × 10⁴/ well) were suspended in 300 μL serum-free medium and seeded in the upper transwell chamber (8 μm pore size, BD Biosciences). For invasion assay, cells in serum-free medium were placed into the upper chamber of an insert coated with Matrigel (BD Biosciences). After incubation for 12 h at 37 °C, non-migrated or non-invaded cells on the upper membrane were removed by a cotton swab. Cells that had migrated or invaded through the membrane were stained with 0.1% crystal violet and three fields were randomly selected for cell number counting.

BALB/c-nude mice (4–5 weeks of age, 18–20 g) were purchased from the Center of Experimental Animal of Third Military Medical University, China and housed in a sterile environment. For tumor formation experiment six 4-week-old female BALB/c-nude mice were randomly divided into two groups (n = 3/group). MDA-MB-231 cells (8× 10⁶) with ALX4 or control vector stably expression were injected subcutaneously into the right flanks of the nude mice, respectively. The tumor volume was determined using the eq. V = 0.5 × D × d² (V, volume; D, longitudinal diameter; d, latitudinal diameter). The developing tumors were observed over the next 33 days. For metastasis experiment six 4-week-old female BALB/c-nude mice were randomly divided into two groups (n = 3/group) and MDA-MB-231 cells (6× 10⁶) with ALX4 or control vector stably expression were injected via tail veins, respectively. Tumor metastasis were observed by H&E staining and qualified by human-specific β₂-MG (beta-2-microglobulin) [23] 5 weeks later. All experimental animal procedures were approved by the Institutional Animal Care and Use Committee of Third Military Medical University, China.

The previously reported promoter region for CTNNB1 (−2760/+27) was amplified and cloned into the pGL3 vector (Promega, Madison, WI, USA) [24]. MDA-MB-231 and MCF-7 cells were transfected with the pGL3-CTNNB1/promoter together with pIRES2-EGFP-ALX4 or control vector and pRTK-Luc (Renilla-TK-luciferase vector, Promega) to normalize the transfection efficiency. 36 h later, the activities of Firefly luciferase and Renilla luciferase were measured using the Dual Luciferase Reporter Assay System (Promega). For TOP and FOP flash assay, MDA-MB-231 and MCF-7 cells were plated into 24-well plates at a concentration of 2.0 × 10⁴ cells per well. Cells were co-transfected with 300 ng of either TOP flash (T-cell factor reporter plasmid) or FOP flash (mutant T-cell factor reporter plasmid) expression plasmids (Millipore, Temecula, CA, USA), and 300 ng of pIRES2-EGFP-ALX4 or control vector and 30 ng pRL-TK. Luciferase activity was measured in triplicate using the fluorescence microplate reader measurement system Varioskan LUX (Thermo Fisher, Waltham, MA, USA) using a Dual-luciferase reporter kit (Promega) as previously described [25].

TCGA dataset (https://​genome-cancer.​soe.​ucsc.​edu/​proj/​site/​xena) was applied to analysis the relationship between ALX4 expression prognostic outcomes of breast cancer patients. Log in the website and click “launch Xena Browser” and in the first section “Select a Study to Explore”, select “TCGA breast cancer (BRCA)” data set in the case of our study. In the second section “Select Your First Variable”, select “Phenotypic” for the data type and “sample type” for the Phenotype. In the third section “Select Your Second Variable”, select “Genomic” for the data type and input the gene name “ALX4” then select Gene expression. When all the parameters are selected we will proceed to the next page and input “Primary tumor” in the filter actions and click “Filter”. Finally click “Kaplan Meier plotter” the relationship between ALX4 expression prognostic outcomes of breast cancer patients will be showed.

The TCGA breast cancer RNAseq (IlluminaHiSeq; N = 1124) data was applied to analyze the expression of ALX4 and genes related to Wnt/β-catenin signaling. The heat maps of gene expression were generated using https://​genome-cancer.​ucsc.​edu/​proj/​site/​hgHeatmap/​. First click “add datasets” on the left corner and the select “TCGA breast cancer RNAseq (IlluminaHiSeq; N = 1124)”. By inputting the gene names the expression heap map will be generated automatically. For statistically analysis, down load the raw data in “.tgz” form and open it with Excel software and analyze the data with relative statistical methods.

The expression status of ALX4 in breast cancer cell lines and normal breast tissues were detected by RT-PCR, qRT-PCR and WB. The results showed that ALX4 was down regulated in breast cancer cell lines compared with the normal breast tissues both on mRNA and protein level (Fig. 1a, b). To further confirm this observation, the expression pattern of ALX4 was subsequently analyzed using The Cancer Genome Atlas (TCGA) database (111 normal breast tissue and 1097 breast cancer tissue). Consistent with our results, ALX4 expression was down regulated in TCGA breast cancer tissues compared with normal breast tissues (P < 0.01) (Fig. 1c). Collectively these results suggested that ALX4 was down regulated in breast cancer, thus it may play an important role in breast cancer development.

Our previous study on lung cancer has identified numbers of CpG islands on the ALX4 promoter region [18] and epigenetically silencing by promoter hyper-methylation have been reported to be a major reason for gene down regulation [25‐27]. To investigate whether the expression of ALX4 was associated with promoter methylation in breast cancer, the methylation status of the ALX4 promoter region was firstly detected by MSP. Complete methylated status was observed in three breast cancer cell lines MDA-MB-231, MCF-7 and T47D (Fig. 1d). We further examined the methylation of clinical samples and the results showed that the methylation rate of the promoter region of ALX4 gene was 69.44% (75/108) in breast cancer tissues of patients, and 0% (0/8) in 8 normal breast tissues (Fig. 1d). To further validate the above results, we used the BSP method to quantitatively detect the methylation status of ALX4 promoter in normal breast tissue and two commonly used breast cancer cell lines. The results showed that CpG island methylation was not detected in the randomly selected normal breast tissue, while significant CpG island methylation was detected in the two breast cancer cell lines MDA-MB-231 and MCF-7 (Fig. 1e). These above results demonstrated that the promoter region of ALX4 exhibited hyper-methylation status in breast cancer. To investigate the relationship between low expression and methylation status of ALX4, we first analyzed the TCGA data using cbioportal website (www.​cbioportal.​com). After downloading the original data, we analyzed the relationship between methylation and expression, and found the expression of ALX4 gene was negatively correlated with the methylation degree (P < 0.00) (Additional file: Fig. S1). In order to further verify that down regulation of ALX4 in breast cancer is related to methylation, breast cancer cell lines were treated with 5-aza-dc, a pharmacological inhibitor of DNA methylation as previously described [9, 28] and the results showed that ALX4 expression was restored (Fig. 1f). These results indicated that promoter methylation is responsible for the down regulation of ALX4 in breast cancer.

To investigate the function of ALX4 in breast cancer, we first transfected MDA-MB-231 and MCF-7 cell lines using control vector and ALX4 expression vector and verified the expression of ALX4 by WB and RT-PCR (Fig. 2a). The effects of ALX4 over expression on cell proliferation and viability were firstly detected. A four days growth curve showed that ALX4 over expression inhibited the proliferation of MDA-MB-231 and MCF-7 cells (Fig. 2b). The suppressive function of ALX4 was further verified by colony formation assay and EDU assay. ALX4 over expression inhibited the colony formation and the EDU positive rate of MDA- MB-231 and MCF-7 cell lines (Fig. 2c, d). Furthermore, we found that ALX4 over expression induced G1 / S blockade of breast cancer cell lines (Fig. 2e). We further investigated the effect of ALX4 on breast cancer cell metastasis. Transwell assays with or without matrix gel showed that ALX4 suppressed the migration and invasion ability of MDA-MB-231 and MCF-7 (Fig. 2f). These data suggested that ALX4 inhibited the proliferation and metastasis of breast cancer cell lines.

To further confirm the role of ALX4 in breast cancer cell growth and metastasis. We knocked down ALX4 expression in the MDA-MB-231-ALX4 and MCF-7-ALX4 stably cell lines with ALX4-miRNA. ALX4 expression was reduced in cells transfected with ALX4-miRNA as shown by WB and RT-PCR (Fig. 3a). Knockdown of ALX4 markedly enhanced cell viability and colony formation ability compared with the control group (Fig. 3b, c). Furthermore, knockdown of ALX4 reversed the inhibitory effect on cell metastasis of MDA-MB-231-ALX4 and MCF7-ALX4 stably cells (Fig. 3d). These knockdown results, together with the above obtained data from ALX4 over expression study suggested that ALX4 might function as a potential tumor suppressor in breast cancer.

To further investigate the tumor suppression function in vivo, nude mice xenograft tumor model was applied to determine the oncogenic role of ALX4 in tumorigenicity of breast cancer cells. MDA-MB-231-ALX4 and vector control stably over expression cells were subcutaneously injected into nude mice to evaluate the effect on tumor formation. The result showed that mice receiving MDA-MB-231-ALX4 stably cells exhibited significantly reduced tumor formation ability compared with vector control cells (Fig. 4A-C). Furthermore, we investigated the effect of ALX4 over expression on breast cancer cells metastasis in vivo via tail veins injection method. H&E staining results showed that ALX4 could suppress the liver metastasis (Fig. 4D) and this observation was further confirmed by qPCR using human-specific β₂-MG (beta-2-microglobulin) [23] with mouse-specific β₂-MG as internal control (Fig. 4E). Collectively, these data indicated that ALX4 had a significant effect on impeding tumor formation and metastasis, supporting ALX4 as a tumor suppressor in breast cancer in vivo.

We next went to explore the possible mechanism by which ALX4 suppressed breast cancer progression. Accumulating evidence suggested that abnormally activated Wnt/β-catenin pathway contributed to the malignant phenotype of breast cancer [29‐31]. Using the TCGA data base, we found that ALX4 expression was positively correlated with the negative regulator of Wnt/β-catenin pathway, such as WIF1 and CTNNBIP1, but negatively correlated with Wnt/β-catenin activated genes, LEF1 and Cyclin D1 (Fig. 5a) (Additional file 1: Figure S2 A, B). In light of this observation, we hypothesized that ALX4 may exerted its inhibitory function by disrupting the Wnt/β-catenin signal. To test this presumption, we first performed Top/Fop flash reporter assay, the results showed that ALX4 could significantly inhibit the transcription activity of β-catenin/T-cell factor (TCF) compared with the vector control group in MDA-MB-231 and MCF-7 cell lines (Fig. 5b). We subsequently analyzed the expression of the downstream target genes of Wnt/β-catenin pathway, and found that ALX4 could decrease the mRNA and protein level of CyclinD1, c-Myc and MMP7 (Fig. 5c, d). Next we thought to figure out the possible mechanism by which ALX4 could interrupt the Wnt/β-catenin signaling. As a transcription factor, we first examine the effect of ALX4 over expression on the transcription of β-catenin (the key signal transmitter involved in Wnt/β-catenin pathway). However, luciferase reporter assay showed that ALX4 had no significant inhibition effects on the transcription of β-catenin in the two breast cancer cell lines (Fig. 6a). Furthermore no significant change was observed for the mRNA expression of β-catenin after ALX4 overexpression (data not shown). Nevertheless we found that the protein level of β-catenin was decreased after ALX4 expression (Fig. 6b). Accumulating evidence indicated that the protein level of β-catenin was tightly controlled by the degradation complex composed by AXIN, ICAT, APC and GSK-3β by which GSK-3β phosphorylates β-catenin leading to its proteolytic degradation [32‐35]. Thus we speculated that ALX4 may promote the phosphorylation degradation of β-catenin. To test this hypothesis, we detected the relative protein expression and the results showed that ALX4 induced the level of p-β-catenin and GSK-3β (Fig. 6b). To further investigate whether GSK3β play important role in ALX4 mediated β-catenin phosphorylation and degradation, we knocked down GSK3β expression in MDA-MB-231-ALX4 stably cell line using siRNA and the level of p-β-catenin was reduced while the protein level of β-catenin was recovered (Fig. 6c). We further consolidated our observation by detecting the expression of ALX4, β-catenin, p-β-catenin and GSK-3β in xenograft tumor tissues by WB (Fig. 6d). Furthermore, by analyzing the TCGA breast cancer data, we found that the expression of ALX4 was positively correlated with the expression of the key members of the “β-catenin degradation complex” (Additional file 1: Figure S2C). To define whether β-catenin was required for the anti-tumorigenesis function of ALX4, we re-expressed β-catenin in MDA-MB-231-ALX4 stably cell line (Fig. 7a). β-catenin re-expression reversed the inhibitory effect of ALX4 on cell proliferation and metastasis (Fig. 7b, c). These data indicated that β-catenin was important for the tumor suppression function of ALX4 in breast cancer. We further examine the effect of GSK3β knock down on the phenotypes of MDA-MB-231-ALX4 stably cell line. Knock down of GSK3β was confirmed by WB (Fig. 7d), and the results showed that knock down of GSK3β could reverse the inhibitory effects of ALX4 on cell proliferation and metastasis at least in part (Fig. 7e, f). Taken together our results supported that ALX4 inhibited the progression of breast cancer through interfering Wnt/β-catenin signaling via promoting phosphorylation degradation of β-catenin in a GSK3β dependent manner (Fig. 7g).

To investigate the clinical significance of ALX4 expression in breast cancer patients, we first conducted IHC on a TMA containing 142 breast cancer patients with overall survival clinical statistics. After IHC, we used the scoring system to consolidate the results for intensity and positive staining percentage. Based on the results, positive staining of tumor cell was quantified and classified into two groups: high and low according to the median score of 142 breast cancer patients (Fig. 8a). Survival analysis using Kaplan-Meier method and log rank test showed that patients with high ALX4 expression (n = 42) presented a longer survival time than that of low ALX4 expression (n = 100) (P < 0.01) (Fig. 8b). To correct for bias caused by univariate analysis, ALX4 expression as well as other parameters were examined in a multivariate Cox-regression analysis (after adjusting for age, histological grade, clinical stage, tumor size and lymph node status). In addition to tumor size (HR = 1.268, P = 0.00), ALX4 expression was found to be an independent prognostic factor (HR = 0.345, P = 0.01) for the overall survival of breast cancer patients (Fig. 8c).

To further confirm these results, we performed a meta-analysis of the association between ALX4 gene expression and prognostic outcomes among 1080 breast cancer patients using the TCGA data base. The results showed that high expression of ALX4 predicted longer survival time (n = 1080, p = 0.00, 2 group; n = 1080, p = 0.03, 3 group) (Additional file 1: Figure S4). Next, we sought to determine the impact of ALX4 expression on patient survival considering tumor size, clinical stage, lymph node and histological grade. After stratifying patients based on ALX4 expression, we analyzed patients’ survival data and found a significant association between ALX4 expression and patients at clinical stages II + III (P = 0.00) (Fig. 8d, e). We further investigated the relationship between ALX4 expression and clinical parameters, and found that ALX4 expression was evidently correlated to clinical stages (n = 140, P = 0.01), tumor size (n = 142, P = 0.01) and lymph node status (n = 138, P = 0.00) of breast cancer patients (Table 1). Taken together our data showed that ALX4 is an independent favorable prognostic factor for breast cancer patients and its expression is associated with clinical parameters.